Floodplain Mapping & LiDAR: Can There be Too Much Data?
Presented by:
Virginia Dadds, P.E., CFM, LEED AP
Aurore Larson, P.E., CFM
Jason Wheatley, GISP
October 21, 2010
LiDAR – the Giant Step Forward
Can there be too many points?
100’x100’ LiDAR Grid
•File size 2 GB
•Frequently corrupted
•Long processing times
•Difficult to append more
data if needed
Bottom Line – Blown Budgets!
Finding a Solution
• Tasked Jason with finding a
better way
– No loss in profile accuracy
– Floodplain lines must pass checks
– Overlay solution on full LiDAR
products for comparison
– Fully document the process for all
users
– Process can be improved as
needed
Inspiration/Problem Identification• Map Production
– Panning/Zooming draw times• Several second refresh rates
• Large vector datasets with excessive detail
– Printing• Larger files with longer print
processing
• Storage/Serving
– Large vector datasets
– LiDAR
– OrthoImagery
Inspiration/Problem Identification• Surface processing
– Buffer waterways to generate “domain”
– Extract LiDAR groundshot from domain
• May not have enough coverage
– Construct TIN ground surface for flood extraction
• Studies in FEMA Region 7 – Iowa
– High-resolution LiDAR point files
(LAS and XYZI) available from the
GeoTREE Iowa LiDAR Mapping
Projecthttp://geotree2.geog.uni.edu/lidar/
Region 7 – Iowa LiDAR (Boone County)• Voluminous data
– 1.4m avg. point spacing
– 2.5 Mil groundshot
points per 4.0 Mil m2
(approx. 1.5 sq. mi.)
– 400 tiles in county
– Approx. 1 Bil groundshot
points in county
• Extremely difficult to process
seamless TIN surfaces for
larger domains
Goals• Improve speed
• Storage savings
• Network performance
• Time savings
– Dedicated to QA/QC
• Effective products
– Information that is optimized for target scale
– More representative of real-world features
Optimization• “Less Is More”
• “Sweet-spotting”
• Improve performance/efficiency of existing and future
processes through generalization of mapping
inputs/outputs
• “Sweetspot” source data to produce most effective
information with least effort
Proposed Solutions• Generalize Product (vector)
– Smoothing/Simplifying lines
• Must meet FEMA DFIRM mapping standards (FBS Audit)
• Still requires TIN generation
• TIN extraction not uniform so process is more difficult
• LiDAR Thinning– Iowa possesses little relief
• Still requires more processing/storage to generate TIN
• Eliminating detail from Raw data
• Raster Elevation Surface– Generate GRID(s) (2m cellsize) from raw groundshot
• Applies point mean to each cell
• Easy to control generalization
• Smaller file size
• County-wide surface
TIN vs. GRID
• Difference in level of detail, or just a difference in interpolation?
• TIN– Elevation of each point is preserved
• Vertical error (+/-7”) also preserved
• Eliminates area from laser pulse(0.5m – 1m)
– Slope/Aspect determined bytriangulating three adjacent points
– Vertices of extraction non-uniformdue to varying triangulations
• Harder to select generalizationtolerances
– Greater uncertainty in samplevoids
TIN vs. GRID
• GRID
– Elevation points are “leveled” through cell averaging
• Vertical error also leveled
– Applies elevation values to anarea rather than specificx/y coordinate
– Vertices of extraction are moreuniform due to equal cell size
• Easier to select generalizationtolerance
– Interpolation considers moreinformation in void areas
TIN 2M GRID
2M 3Cell 2M 11Cell
TIN vs. GRID 3c
1:300 scale
Surface Tests
• TIN from raw LAS extraction (groundshot)
• GRID (2m cellsize)
• GRID Re-sampled (Mean Neighborhood-square)
– 3 cells
– 5 cells
– 7 cells
– 11 cells
• 3 Water features
– Des Moines River (large), North River (med.), and Butcher
Creek (small)
TIN GRID 2m GRID 2m3c
GRID 2m5c GRID 2m7c GRID 2m11c
1:2,000
Surface Selection
• 3cell Re-sampled GRID surface
– Smooth, cartographic quality delineation
– “Clean” at 1:6,000 scale
– Upheld accuracy standards
• FBS Audit
– Two pass test
• Pass 1 - Line position compared to source models (<= 1’)
• Pass 2 - Line must fall within 38’ of the elevation match
GRID 2m3c
FBS Audit results - Des Moines River
Source
Surface
Audit
Surface
Water
SurfaceSample Size
Max/
Average
Difference
Pass 1 - %Pass 2 - %
(38ft)
Pass 3 - %
(25ft)
Pass 4 - %
(5ft)
TIN TIN TIN 491 4.54’/0.80’ 68.64% 100% 100% 96.33%
GRID 2m GRID 2m GRID 2m 491 3.72’/0.43’ 88.80% 100% 100% 98.98%
GRID 2m 3c GRID 2m GRID 2m 439 2.59’/0.44’ 89.29% 100% 100% 99.09%
GRID 2m 5c GRID 2m GRID 2m 412 3.99’/0.70’ 74.03% 100% 100% 93.20%
GRID 2m 7c GRID 2m GRID 2m 387 4.91’/0.91’ 65.63% 100% 100% 87.08%
GRID 2m 11c GRID 2m GRID 2m 336 9.02’/1.40’ 51.79% 100% 99.70% 70.83%
GRID 2m 3c TIN TIN 439 2.77’/0.51’ 86.10% 100% 100% 97.69%
FBS Audit results - North River
Source
Surface
Audit
Surface
Water
SurfaceSample Size
Max/
Average
Difference
Pass 1 - %Pass 2 - %
(38ft)
Pass 3 - %
(25ft)
Pass 4 - %
(5ft)
TIN TIN TIN 1084 11.55’/1.12’ 66.88% 100% 99.72% 83.39%
GRID 2m GRID 2m GRID 2m 1068 4.02’/0.36’ 91.57% 100% 100% 98.13%
GRID 2m 3c GRID 2m GRID 2m 991 5.39’/0.41’ 88.80% 100% 100% 96.57%
GRID 2m 5c GRID 2m GRID 2m 913 4.85’/0.54’ 83.46% 100% 100% 93.10%
GRID 2m 7c GRID 2m GRID 2m 782 7.28’/0.67’ 79.67% 100% 100% 87.98%
GRID 2m 11c GRID 2m GRID 2m 604 5.46’/0.99’ 65.07% 99.17% 99.01% 70.53%
GRID 2m 3c TIN TIN 991 6.34’/0.46’ 85.77% 100% 100% 96.37%
FBS Audit results - Butcher Creek
Source
Surface
Audit
Surface
Water
SurfaceSample Size
Max/
Average
Difference
Pass 1 - %Pass 2 - %
(38ft)
Pass 3 - %
(25ft)
Pass 4 - %
(5ft)
TIN TIN TIN 521 4.75’/0.29’ 95.20% 99.81% 99.42% 98.08%
GRID 2m GRID 2m GRID 2m 484 4.33’/0.37’ 90.91% 99.59% 99.17% 95.45%
GRID 2m 3c GRID 2m GRID 2m 441 4.27’/0.43’ 87.53% 99.77% 98.87% 95.69%
GRID 2m 5c GRID 2m GRID 2m 409 4.79’/0.51’ 86.06% 99.76% 99.27% 91.93%
GRID 2m 7c GRID 2m GRID 2m 386 6.00’/0.6’8 78.76% 99.74% 99.48% 85.23%
GRID 2m 11c GRID 2m GRID 2m 370 7.07’/0.99’ 67.30% 98.92% 97.57% 70.27%
GRID 2m 3c TIN TIN 441 4.69’/0.50’ 86.85% 99.77% 99.32% 95.01%
Surface Processing Comparison
SurfaceSpatial
Extent
Overall
Time
Direct
Labor Time
File Size
(rounded)MB/sq.mi. Comments
TINNorth River
(4 sq. mi.)12 hours 9 hours 400 MB 100 MB
• Large
footprint
•Extensive
staff time
GRID 2m
(and 4
versions)
Warren
County
(715 sq. mi.)
4 hours 1 hour 1800 MB 2.5 MB
•Smaller
footprint
•Simple
processing
Line Generalization
Location
Line
Length
Pre-Simp
# of
Vertices
Pre-Simp
Line Length
Post-Simp
# of Vertices
Post-Simp
Line
Length
Reduction
%
Vertex
Reduction
%
North River35,118 m/
115,217 ft3,763
34,440 m/
112,992 ft2,482 2% > 34%
Des Moines
River
14,786 m/
48,510 ft2,884
14,611 m/
47,936 ft1418 1% > 51%
Butcher
Creek
14,935 m/
48,999 ft2,385
14,461 m/
47,445 ft1,547 3% > 35%
Line Generalization Results
• TINs used for re-delineated flooding
• Flooding produced 1,280,003 vertices
• Simplified by 1m = 177,311 vertices
• Poly size 39.1MB vs. 5.5MB
1:6,000
1:500
Generalized FBS Audit results
Location Sample SizeAverage
DifferencePass 1 - % Pass 2 - % (38ft) Pass 2 - % (25ft) Pass 2 - % (5ft)
North River
(Pre-Simp)991 5.39’/0.41’ 88.80% 100% 100% 96.57%
North River
(Post-Simp)989 5.39’/0.41’ 88.88% 100% 100% 96.26%
Des Moines
River
(Pre-Simp)
439 2.59’/0.44’ 89.29% 100% 100% 99.09%
Des Moines
River
(Post-Simp)
432 2.68/0.50’ 86.11% 100% 100% 98.38%
Butcher Creek
(Pre-Simp)441 4.27’/0.43’ 87.53% 99.77% 98.87% 95.69%
Butcher Creek
(Post-Simp)425 4.11’/0.44’ 88.71% 99.76% 99.53% 95.06%
Conclusions• LiDAR elevation data for Iowa could afford
generalization
• Time savings allows for more time dedicated to
QA/QC
• Produce quality product more efficiently
• TINs are not necessarily more accurate than Rasters
when interpolating surfaces
• Capable of meeting FEMA DFIRM mapping
specifications
Benefits Realized• Surface generation was completed in 1/3rd of the time
it takes to produce TIN surfaces
• Estimated 97% storage savings
• Linework more smooth, representative of real world
phenomena, and streamlined map production
Comments/Questions?• Acknowledgements
– Aurore Larson, P.E., CFM – Greenhorne & O’Mara – Water Resources Services
– Carmen Burducea, CFM – Greenhorne & O’Mara – Water Resources Services
– Zachary J. Baccala, Senior GIS Analyst – PBS&J – Floodplain Management Division
• References– Cohen, Chelsea, The Impact of Surface Data Accuracy on Floodplain Mapping,
University of Texas, 2007.
– Foote, K.E., Huebner, D.J., Error, Accuracy, and Precision, The Geographer’s Craft Project, Dept. of Geography, The University of Colorado at Boulder, 1995.
– Galanda, Martin, Optimization Techniques for Polygon Generalization, Dept. of Geography, University of Zurich, 2001.
– Lagrange, Muller, and Weibel, GIS and Generalization: Methodology and Practice, 1995.
– Li, B., Wilkinson, G. G., Khaddaj, S., Cell-based Model For GIS Generalization, Kingston University, 2001.
– North Carolina Floodplain Mapping Program, LiDAR and Digital Elevation Data (Factsheet), 2003.